a ; <¢ ;

Pay hdres

oo vi

eae Poe eee

ee as ee ee ey

' = 2

0

: rf al) BAM myn : ny

f

wea nie! ea. ap

> ip 7 tal

Pr by it px : mo Ue wn

re A

Ps es ag <7 »

ae. Bs PROCEEDINGS <<

OF THE

AMERICAN PHILOSOPHICAL SOCIETY

HELD AT PHILADELPHIA

FOR

PROMOTING USEFUL KNOWLEDGE.

VOL. XXXIX.

JANUARY TO DECEMBER, 1900.

PHILADELPHIA : THE AMERICAN PHILOSOPHICAL SOCIETY. 1900.

PROCEEDINGS

OF THE

AMERICAN PHILOSOPHICAL SOCIETY HELD AT PHILADELPHIS FOR PROMOTING USEFUL KNOWLEDGE.

January, 1900. ‘No. 161:

VoL. XXXIX.

Stated Meeting, January 5, 1900. Vice-President SELLERS in the Chair. Present, 20 members.

Messrs. Daniel Baugh, Leslie W. Miller, Henry Kraemer, Coleman Sellers, Jr., Edwin Swift Balch, Lindley M. Keas- bey, Francis Rawle and Marion D. Learned, newly elected members, were presented to the Chair and took their seats in the Society.

Letters acknowledging election and accepting membership were read from Messrs. Edwin Swift Balch, Edward Coles, Coleman Sellers, Jr. Thomas Harvey Dougherty, Henry Kraemer, Alexander Jay Wurts, Albert Matthews, Lindley M. Keasbey, J. Rodman Paul, Daniel Baugh, Leshe W. Mil- ler, Thomas Corwin Mendenhall, Otis Tufton Mason, J. Dun- das Lippincott and Robert C. H. Brock.

The list of donations to the Library was presented and thanks were ordered therefor.

The decease of the following members was announced :

Dr. Elliott Coues, of Washington, D. C., at Baltimore, on December 25, 1899, in his 57th year.

Sir James Paget, Bart., of London, on December 30, 1899, aged 85 years.

Mr. A. Radcliffe Grote presented a paper entitled, “ The Descent of the Pierids.”’

The Judges of the annual election for Officers and Council-

PROC. AMER. PHILOS. SOC. XXxrx. 161. A. PRINTED MARCH 15, 1900.

4 GROTE—THE DESCENT OF THE PIERIDS. [Jan. 5,

lors, held this day between the hours of two and five in the afternoon, reported that the following-named persons had been chosen according to the laws, regulations and ordinances of the Society to the offices for the ensuing year: President. ; Frederick Fraley. Vice-Presidents. Coleman Sellers, Isaac J. Wistar, George F. Barker. Secretaries. I. Minis Hays, Frederick Prime, Samuel P. Sadtler, Richard A. Cleemann. Treasurer. Horace Jayne. Curators. J. Cheston Morris, Benjamin Smith Lyman, Henry Pettit. Councillors to serve for three years. George R. Morehouse, Patterson DuBois, Henry C. Trumbull, C. L. Doolittle.

The Society was adjourned by the presiding officer.

THE DESCENT OF THE PIERIDS. (Plates I-IV.)

BY A.-RADCLIFFE GROTE, A.M.,

Vice-President of the American Association for the Advancement of Science, Etc. (Read January 5, 1900.)

GENERAL WING CHARACTERS OF THE PIERIDS.

A horde of bright butterflies with entire wings, simply marked in gay and contrasting colors: white, yellow, orange, red, brown- ish and black, rarely with a blue reflection or a lilac shade, but often marbled beneath with lichen green, of the most general dis- tribution, and numbering almost all of the very few species of diurnals counted as injurious to vegetation. The neuration of the

1900. } GROTE—THE DESCENT OF THE PIERIDS. 3)

Pierids may be best compared with that of the Limnadidz among the brush-footed butterflies, but is most clearly more specialized. The cells are closed, but the transverse vein is often degenerate. There are still, sometimes, very small backward spurs, slight remainders of the vanished base of the median system. The cubital cross-vein has left no mark. The fork to the second anal vein of the fore wings at base is very thin and transparent. The trace of the first anal vein has usually disappeared. The upper me- dian branches ascend the radius, in the specializing movement attending the disintegration of the system of the media. The radial branches are reduced at times to three, the usual number is four, and very rarely the five original branchlets are retained on the primary wing. On the hind wings the humeral cell is reduced, often the central slit can with difficulty be discerned, and sometimes the veins appear to be here completely fused and the subcostal to be single. The humeral spur is usually present, but vanishes in spe- cialization. ‘There are two anal veins on the secondaries. These characters show that the Pierids belong to the second division of the diurnals, in which the free and short downwardly curved third anal vein of the fore wings of the Papilonides is absent, viz., the Hesperiades.

_An antecedent specialization has taken place on the hind wings in that the radius is single, or, if we adopt the view that it is fur- cate at base and that the costal cell is formed by Rr, two branched. Present specialization seems still to favor the seconda- ries, as in those Nymphalids in which the cell is open on hind wings and closed on primaries, the reverse‘appears never to happen. But in the Parnassiidz and the Pierids the progressing disintegration of the median system appears more marked now on the primaries, where the upper branches ascend the radius and the cross-vein weakens.

PHYLOGENY.

Not without much trouble have I been able to detach the ter- minal branch of the Hesperiades from what clings, and yet seems not to belong to it, on the lepidopterous tree. From a common pattern with separated veins, the middle branch of the disinte- grating median system fading out at length (2% sztz) in both Blues and Skippers, owing to the shrinkage of the connecting cross-vein, the wings of the Lyczenids have evidently emerged, losing the gen-

6 GROTE—THE DESCENT OF THE PIERIDS. (Jan. 5,

eralized Hesperian number of radial veins in the operation. The principal feature of the Lyczenid-Hesperid offshoot is nearly real- ized by the next departure, that of the Nymphalids, in the ‘ long- fork’’ of the Charaxine. Finally, we find the main branch itself culminating in the Pierids, which, while otherwise agreeing well with the brush-footed butterflies in neuration, are to-day nearly all relatively more specialized, the radius of the primary wings being usually only four-veined, the upper veinlets of the media tending also to travel up the vein and arise from beyond the cell. From this branch of the Hesperiades I have tried to disentangle the over- lapping Papilionides, with their short, free and downwardly curved third anal vein on the fore wings, the second anal vein showing no thinner basal fork, as it does in the Hesperiades and most of the rest of the Lepidoptera. Even if, as appears to be Mr. Quail’s opinion, I have not succeeded in getting rid of this Papilionid spray alto- gether, I have at least shown that its supposed issue from the Hes- periades branch, between the Blues and the Skippers, affords no growing point, the two latter being consecutive groups. And this was my main endeavor, to take Paf/io out from between Lycena and Hesperia.

And now having cleared as well as I am able the branch down to the Skippers or Hesperiadee, let me try to get lower still and find the connection of the Pieri-Hesperiadz with the trunk of the lepidopterous tree, tne Tineides. And here I am met by an ob- struction, and also a general prepossession out of which this ob- struction arises. I will try to deal first with the prevalent assump- tion, which is, that the diurnals outrank the rest of the Lepidoptera, are more specialized, and that they represent, so to speak, the out- come of the evolution of the order, which has travailed long to produce them. This view is pictorially represented by all the genealogical trees I have seen, and we may take that of Dr. Pack- ard (1895) as an example. Here the butterflies unfold themselves in the left-hand corner, at the top of the page, and the support- ing twigs and limbs are supplied by all the moths, arranged more or less after the catalogue sequence, until we get down to the roots, represented by the Tineina and Eriocephalide. How it comes, after such a progression, that the pupz of the diurnals still show Tineid features, as evidenced by Dr. Chapman, how it comes that the wings of butterflies retain generalized structures as I try to prove, seems, and no doubt is, incompatible with this perfect

“I

1900. | GROTE—THE DESCENT OF THE PIERIDS.

scheme. Let us first take up the wings of some moths and show by comparison that they may be more specialized than those of the butterflies.

Specialization is shown by the different groups of the diurnals in a variety of ways and directions. It may be said of the Papilio- nides, that their main advance is in the reduction of the anal veins of hind wings and the hollowing out of the internal margin of the wing, as shown in an extreme degree by Parnassius. Now, when we examine the wings of the Saturnians, we see, in such genera as Rothschildia, Samia and Callosamia, the number of anal veins equally reduced to one, the internal margin equally hollowed out. It may be said again of the Pierids, that their main advance is shown by the diminution of the branches of the radius of the fore wings, which from five in number are reduced to four and three. Turning back to Rothschildia, Samia, Attacus, etc., we find the same reduction. It may be said in turn of the Nymphalids, that their main advance is shown by the opening of the discal cell, the more complete disintegration of the median system of neuration. Again in the same genera of Attacinee, we find the cell even more completely open, the cross-vein vanished, the centre of the wing longitudinally clean from veins, the median system completely divi- ded between and absorbed by the radius and cubitus (see my figures of Saturnian wings: Bestrag zur Classification der Schmet- terlinge). Thus the wings of the single Attacid genus Rothschit- dia combine in a high degree the main features of the specializa- tions shown separately by three most important groups of the diurnals. The Attacid wing is far more specialized than the wing of any butterfly. Not only this, it has also lost all residual features which are retained by the butterflies. Let us enumerate some of these: The backward spurs, traces of the base of the median sys- tem, to the discal cross-vein, present in many butterflies, are lost. All trace of the cubital cross-vein has vanished from the Attacid primary. This is strong in Pago and allies, quite marked in a number of Nymphalid forms, Morphide, Limnadidez, Heliconiide. On the hind wings, the humeral cell, formed by the forking of the radius, closed by the residuary vein Ri, has also gone. This is found, in various stages of completeness, in many butterflies. The humeral spur (precostal spur) of the butterflies has disappeared in the Saturnians, while the frenulum has departed in both groups. If we take the specializations as indicating younger forms, then the

8 GROTE—THE DESCENT OF THE PIERIDS. [Jan. 5,

Attacid genera are more modern than the diurnals, which they thus clearly outrank. Now the indications of the base of the median system, the presence of cross-veins, the forking of the radius of hind wings, are primitive or Tineid characters, and the conclusion is here again irresistible, the diurnals are nearer to the Tineid trunk of the lepidopterous tree than the Attacinz, because they retain more generalized features in the neuration, I think these considerations should divert the general prepossession that the butterflies are the final product of the lepidopterous tree, or that this corresponds in reality with Dr. Packard’s plan, which, with all allowances, appears to be still governed by this idea.

The obstruction to proceeding further with the stem of the Hes- periades arises from that extensive group which Dr. Dyar has dis- cussed under the name Bombycides (Agrotides), Proc. Bost. Soc. NV. H., 27, 127-147, 1896. I do not think sufficient attention has been paid to Dr. Dyar’s separation of the higher moths into dis- tinct phyletic groups upon the characters offered by the position of the larval tubercles. If these characters hold, and the groups es- tablished ‘by Dr. Dyar are separate and monophyletic branches of the lepidopterous tree, an immense stride toward our comprehen- sion of the order will have been made. The neuration offers, in the Sphingides, Saturniades, Bombycides, no exclusive character, nothing to absolutely limit these groups or to satisfactorily distin- guish them from the Hesperiades. These latter are open to the moths generally, but, if Dr. Dyar’s characters hold, I can now sweep aside this vast branch of the Bombycides and account for correspondencies by the fact that the branches have the same origin, but an independent and parallel growth. We need not pass through the Bombycides on the way to the root of the Hesperiades, nor could we very well, seeing they are now about equally specialized. In the same way we may neglect the Sphingides and Saturniades. Arriving in this way at the Tineides, we find in the Tortricidz a family with partially concealed larve, in neuration and wing form having marked affinities with the Skippers. Beyond these are the related Cosside, with concealed, grublike larvz, the neuration generalized in that the median system is complete. Both these families have the thinner basal loop to the second anal vein of primaries in common with other moths. While there may be no positive proof, it seems to me probable that these families of Tinei-

1900. ] GROTE—THE DESCENT OF THE PIERIDS. 9

des have been developed from the same phyletic stem which finally produced the Pierids and Hesperiades.

If the alternative conclusion be adopted which I have suggested, that the so-called ‘‘ precostal spur ’’ is really homologous with the ‘‘ humeral veins’’ of the Lachneids, we must seek for the origin of the Hesperiades in common with that branch of Dr. Dyar’s Bom- bycides in which it is developed, and probably simultaneously from antecedent Tineid types. This would seem to weaken the prob- ability, suggested by Sir George Hampson’s phylogeny, that they are derived from forms which have not developed the structure, unless we assume the humeral spurs to be characters of convergence. This humeral vein, which I shall call in the progress of this paper the ‘‘ humeral spur,’’ is single in the Hesperiades, and may be here a survival, others having perhaps disappeared. This survival itself gradually fades away ; in Zervzas it isa mere knob, in WVatha/is it has nearly vanished. ~

The primitive Hesperiad had probably a wing in which the veins and branches were all separate, the base of the median sys- tem more or less fully developed and with the first anal existing. The second median was central. The cubital cross-vein was present. The fork of the second anal was possibly expanded into a longitu- dinal vein connecting with second anal by across-vein. The prim- itive Papilionid was probably propinquitous, having most features in common, but the third anal was longer and unconnected, and there are indications from the position of the second median that it was not’ central as in the primitive Hesperiad from the different shape of the discal cell. On the hind wings the primitive Hespe- riad would have lost the three outer branches of the radius, but the first would be present, closing the subcostal cell as in the Sphingi- dz. On the costal region of primaries the costal vein might have been functional, but it had probably already left the secondaries. Instead there might have been a greater development of humeral spurs.

It will accord best with the paleontological record if we assume an early and comparatively quick evolution for the diurnals, and as in these are included the Papilonides we must finally say a word as to their presumed origin.

I have recorded elsewhere my search for the short downwardly curved third anal vein of the Papilionides’ primary wing. Mr. Meyrick having figured it in the Geometridz, I sought for but

10 GROTE—THE DESCENT OF THE PIERIDS. Jan. 9,

could not verify its existence. Indeed it is quite clear that Mr. Meyrick’s figure of the wing of Venilia macularia is impossible. For supposing that we could homologize the loop of the second anal with the third anal of the Papilionides, Mr. Meyrick’s figure would depict a moth with four anal veins to the primary, since both loop and downwardly curved vein are given. It seems as though the fork or loop to the second anal vein must exclude the third anal vein of the Papilionides and should therefore be homo- logous with it. In the Saturnian genera Ze/ea and Actas the loop is continued as a short spur. This looks as though the loop might be the remains of a longitudinal third anal vein, which has been joined to the second anal bya cross-vein, while the external portion has afterwards degenerated.

This view is supported by the fact that the prolongations to the pure and simple loop appear in the Tineid families Cossidz and Psy- chide, but, in either case, the diphyletism of the diurnals is un- touched, since we can reach the Cosszde without again meeting the free and downwardly curved third anal vein of the Papilionides. The third anal would have been merely connected with the second in one line and not in the other. At this moment the Papilionides hang in the air and their ancestry is undiscovered. | We may sup- pose them to have been evolved at the same geological period with the Hesperiades, whose origin we can trace, and that the record has in their case become lost.

ABERRANT TYPES.

Having thus indicated the course probably taken by the ances- tors of the two divisions of the butterflies (Lepidoptera which of old loved light rather than darkness and hence may have become what they are), we can turn to a brief consideration of three dis- sonant minor groups which hang upon the skirts of the Pierids without any traceable connection. These are the Leptidiane, the Dismorphiane and what we must call the Pseudopontiade, for, according to a note in the Hxtomologist’s Record, Mr. W. F. Kirby has shown that Psewdopontia of Plétz antedates by a little Gonophle- bta of Felder, and Mr. Scudder’s statement that. we should prefer the latter is thus amended. The two former groups seem closely allied by the extraordinary character of the cubital movement of the middle vein of the median series. The Leptidiane are pecu-

1900. } GROTE—THE DESCENT OF THE PIERIDS. ist

liar in the absence of the fork to the second anal vein of the fore wings. They are thus comparatively specialized, although the ra- dius retains the five branchlets. —The common characters of the two groups with their five-branched radius and elongate wings apparently warrant a phylogenetic association, though in their special features sufficiently distinct. The species are, technically speaking, ‘‘ long wings,’’ forms in which the primaries have the branches of the radius removed outwardly, sometimes, as here, entirely beyond the cross-vein of the discal cell. The passage of the ordinary broad- winged type into that of the “‘ long wings ’’ I have recorded in the Nymphalids. Since I find no connection here of the two types of wing I have separated the Dismorphiadz from the Pierids, and ex- clude them when using in this paper the latter term.

The single species and type of the Pseudopontiade may be called, in contradistinction perhaps to almost all other Lepidoptera, a ‘‘round wing.’’ In this place it is sufficient to state that the neuration is extraordinarily specialized. The radius is only three- branched and the median series shows the Pierid movement in specialization, the two upper branches both actually ascending the radius. But on the hind wings three anal veins have been pre- served, and the origin of this isolated type must be sought for at a period when this structure prevailed. It may have survived in this instance owing to the necessity for sustaining the spherical shape of the wings. We may now connect this form directly with Tine- ides which have the same three anal veins on tbe hind wings pres- ents) Consult figure, Proc. Aw. Prin. Soc, xxxvil, Pl. 1) Fie. 7.

12 GROTE—THE DESCENT OF THE PIERIDS.: (Jan. 5,

PROBABLE PHYLOGENY OF THE HESPERIADES.

& g2) _ w w re) Onis ee & Ke} co) -_ _ iB = ° x S en R | oS ow 35 o } Oy | | wn | = : 4 | | s | | fey | & | .S) | } ES 4 Bat K 5: F ¥ | 8 = | fe oC plc} | a, | & ise & Z, | | cc ————— ie} as i Ay —=— — —_ = ae . — R 3 a) ‘S i a, ~ “ = S & SI w | 3 A ae) ese vo w Qu : : ; Wis + ° o z a] HS 8 S S = = zs 5 a wn © S ro¥y S 5 as n if) oN = 8 ra) ES 2 2 5 os] 12 “s © Aa m4 ja = G CLASSIFICATION.

The characters used in the following table are not exclusive, but serve to distinguish the material here discussed :

Hind wings with a single anal vein...... .... »PAPILIONIDES.

1900. j GROTE—THE DESCENT OF THE PIERIDS. 15

Hind wings with two anal veins............... HESPERIADES. Radius of fore wings 3 to5 branched; first radial branch fromeapove: cell: 4Z2 uradialls... acaws os he ow 0 PIERIDIDA.

Radius of fore wings 5 branched ; all radial branches arising beyond cell; 472 central or cubital... . DISMORPHIADA. Discal cells not reduced; the cross-vein concave ; J/2

cubital’ on fore Wings.!.05. 0! ede te Dismorphiane.

Discal cells reduced; the cross-vein convex; J/2 cen-

PRAM ONELOTE. WINGS af Bia alae ates whee. ale Sarde Leptidiane.

Hind wings with three anal veins.............. PSEUDOPONTIAD&.

Family DisMoRPHIAD&.

Wings entire; fore wings with the radius 5 branched, all the short branchlets arising equidistantly beyond the cell. Cells closed ; middle branch of the median series central or cubital. I find only doubtful traces of the fork to second anal vein of prima- ries at base. In the Dismorphianz the male primary is narrow ; hind wings sometimes much enlarged inferiorly ; no trace of costal cell; middle branch of median series cubital ; two anal veins.

Subfamily ZLeptidiane Grote, 1897.

Wings entire, elongate, proportionate. Middle branch of median series central on primaries, cubital on hind wings. Cells closed, reduced to basal third of wings ; cross-vein outwardly convex. No trace of fork to second anal on primaries. Secondaries not enlarged, proportionate.

Outer margin of fore wings acuminate between first and second median branchlets, second and third median propinquitous

Sub DASE FOI CLOSS-VEIIN. not sis aie oe Sie es oe mys olen w/e zalats. Outer margin of fore wings not acuminate ; second and third median! branchlets separate at base.........-........ Leptidia.

Azalats n. g. Type: Leucophasia gigantea Leech, Ent., 23, 46, 1890. Azalats gigantea.

A more specialized form than Zeftdia, as shown by the approxi- mation of the second and third median branchlets on hind wings, the two branches arising close together from the point of convexity of the cross-vein. This insect is described and figured in the But terflies of China, p. 484, Pl. 36, Figs. 10, 11, 1893.

14 GROTE—THE DESCENT OF THE PIERIDS. (Jan. 5,

Gen. Leptidia Bilberg, 1820. Type: L. sinapis. Syn. Leucophasia Stephens, 1827. Leptidia sinapis.

This common European species differs from Aza/ais by the characters given in the above table. From the neuration the genus Leptosia, with the type given by Scudder, does not belong here. In the Handbook of British Lepidoptera, Mr. Meyrick places Lepitdia in the Pierididz, between Luchloe and Pieris.

The wings of the Leptidianze show apparently no sexual character in size and shape, whereas, in the Dismorphianz, the males have the primaries narrowed and the secondaries disproportionately widened.

Subfamily Dzsmorphiane.

This subfamily is closely related to the,preceding, but differs by the sexual disproportion of the wings in the males of the genera allied to Dismorphia, and by the longer and narrow discal cells. I can find no traces of the fork to the second anal vein of fore wings at base except doubtfully in Acmepteron. ‘The Dismorphianze agree in the cubital position of the middle median branch, but this posi- tion is assumed on doth wings, hence it would be herein a more spe- cialized group than the Leptidianz. The parallel between these two groups and the Heliconiide is shown by the neuration, in that the radial branches tend also in the latter family to arise beyond the cell, Neither in this group nor the Leptidianz are there any traces of the cubital cross-vein, hence they are more specialized forms than the Heliconiidz. It is possible the coincidences indicate that they had a common origin, but against this view the fact of the cubital position of the second median branch, as contrasted with its radial position in the Heliconiidz and the brush-footed butter- flies, will militate. The discal cross-vein is, however, concave on fore wings as in the Heliconians, but since it is convex outwardly in the Leptidianz, which otherwise so well agree, the character will not have much weight. The broadened hind wings of male Dz¢s- morphia recall in shape those of Pseudopontia.

The Dismorphianz are remarkable in the cubital position of the middle median branch of fore wings, and this position within the Hesperiades is only repeated apparently in certain Giant Skippers, Megathymidz, which I have had no opportunity to study closely.

1900. ] GROTE—-THE DESCENT OF THE PIERIDS. 15

A trace of this tendency is, however, visible in the Skipper Angiades sylvanus. Here, on the fore wings, this vein, 1/2, still keeping central, has succeeded in sustaining its connection with the cubitus by a true vein ; that is, the lower part of the cross-vein between d/2 and 4/3 has retained its functional character, while the upper portion, between J/2 and J/1, has yielded to disin- tegration and become a scar. Hence J/2 is still fed from the cubitus. On the hind wings vein J/2 has failed to retain the connection, the source of blood supply has given out with the dis- integration of the entire cross-vein, and, in consequence, this mid- dle branch of the media has almost entirely vanished from the surface of the wing.

Of the genera of the Dismorphianz, Mr. Scudder gives the type of Dismorphia : lata ; of Leptalis : astynome. Additional genera are given by Dr. Dixey in a footnote, Zrans. Ent. Soc. London, 1896, p. 65. Mr. W. F. Kirby kindly informs that the types of these are: of Pseudopieris: nehemia ; of Enantia: melite; of Mos- choneura: methymna. Of these types I have been unable to pro- cure specimens of the first and last for study.

Gen. Leftafis Dalman, 1823.

Type: L. astynome. Leptalis astynome.

Radius five-branched, all the branches arising beyond the cell ; #1 fusing above with subcostal vein. Cell closed, long and nar- row ; cross-vein deeply concave, somewhat weakened. First median branchlet arising at junction of cross-vein with radius; J/2 cubital. No trace of fork to second anal. Hind wings with hardly abbrevi- ate somewhat pointed costal, curving toward base of wing; no trace of subcostal cell. First median branchlet has left cross-vein and arises from radius at about one-third of the length from cross- vein to external margin. Cell closed ; J/2 cubital.

This genus is, perhaps, less specialized than Pseudopieris, in which the discal cell is superiorly retreated, leaving //1 to spring from radius beyond it. It is more specialized than xantia (to which it stands nearer) and Acmepieron, because in these the first median branchlet still remains on the cross-vein.

Leptalis melita.

I describe this since the species has become well known, owing

to its having been cited by Fritz Miiller as an example of the primi-

16 GROTE —THE DESCENT OF THE PIERIDS. [Jan. 5,

tive type of coloration of the genus. It is very close to astynome, but represents a slightly more generalized type and must be considered thus a more primitive form. It differs by the minute character that the first median branchlet appears as a continuation of the cross- vein from its upper corner; in astynome the fusion of the upper wall of the vein with the radius can be made out. The radial branchlets are a little longer, but Ar equally fuses with the sub- costa. On hind wings /t ascends the radius for a shorter dis- tance, lagging thus behind astyzome and much as in £xantia. Since I am not in a position to compare Dismorphia, I cannot con- clude as to the absolutely right place for me/a and leave it under Lepialis. Color and pattern do not move necessarily with neura- tion and at times (as in Cumizza hirlanda) are much at variance. Leptalis melia is a more specialized insect, so far as the neuration is concerned, than those forms of Dismorphia, now to be described, in which J7/r still clings to cross-vein before the upper angle of cell. From its position with regard to astynome, it might have served as a palette upon which the heliconine pattern of the wing was subsequently traced, but I think not, the differences in neura- tion being so very slightly to its disadvantage. The costal humeral spur of hind wings is as in astynome. L. melia appears as a natural specialization of Hnantia milete.

I will first take up a form which stands a little apart from the others here treated of :

Gen. Pseudopierts God. and Salv. Type: P. nehemia.

Pseudopterts nehemia (Boisd.).

Radius five-branched, all the branches arising beyond the cell. Fi not fusing above with subcostal vein, but free. Discal cells closed, but curiously retreating superiorly on both wings ; the cross- vein here falling back toward base of wings and leaving JM/1 stranded and fused with radius, at a point opposite 472, so that it has made no independent radial progress. On hind wings the humeral spur is straight, truncate, terminating in a Y-shaped pro- jection.

This form differs from all others examined and the genus is undoubtedly valid, while the rest hold more or less together as phases in a progressive common specialization of the neuration.

lord

1900.) GROTE—THE DESCENT OF THE PIERIDS. Li

Gen. Hnantia Hiibner, 1816.

Type: E.. melite: Lnantia melite (L.).

Radius five-branched, all branches arising beyond cell, but com- pared with Zepfals, the first radial is less advanced, much nearer cross-vein, hence in a more generalized condition. 1, angulate, fusing with subcostal. Cell closed, concave; 4/1, from cross-vein close to radius; J/2, cubital. Hind wings with hardly abbreviate, somewhat pointed costal, turning well to base of wing; no trace of subcostal cell. First median branchlet has left cross-vein and arises from radius at somewhat less than one-third of its length from cross- vein to external margin. Cell closed; JZ2 cubital.

This form is more generalized than Leptalis. Mr. Scudder takes Zicinia as the type of Lxantia, but, for several reasons, I follow Mr. Kirby here.

Gen. Acmepteron God. and Salv. Type: A. nemesis.

Acmepteron nemests (Latr.).

Radius five-branched, all the branches arising beyond discal cell; Az fusing with subcostal vein. Cell closed ; cross-vein me- dially concave ; AZ arising much before upper angle from cross- vein, hence this genus is more generalized than Euantia; Ma, cubital; from the concave portion of cross-vein a strong back- ward spur arises, the relic of base of the median system. Hind wings with humeral spur straight, truncate; cell closed, with Mi arising at upper extremity of cross-vein, hence much general- ized.

This genus comes nearest to the primitive type of the subfamily, being more generalized than any examined by me. There isa dif- ference in the degree of sinuosity of the second anal vein, which is much bent in Acmepteron, but which I have not been able to pre- cise. In this form also I believe to have detected a very faint linear scar at base of second anal, and which I cannot make out in the other genera of the family. It would occur naturally in a general- ized form, but, as I have said elsewhere, the detection of the scar, when faded, is quite difficult and depends on the light in which the tegument is viewed.

PROC. AMER. PHILOS. SOC, XxxIX. 161. B. PRINTED MARCH 18, 1900.

18 GROTE—THE DESCENT OF THE PIERIDS. [Jan, 9,

CONCLUSIONS ON THE DISMORPHIADZ.

This small and peculiar family is of the utmost importance as bringing evidence of the solidarity of the Hesperiades as classified by me. The generalized features are distinctly those of the Skip- pers. The five-branched radius has the branches all equidistant, and they distinctly recall, in their position and course, the branches of the Lycznide and Hesperide. The radial branchlets have moved along; #4 has crept up #5 till it becomes a short fork. The cubital position of the second median branchlet shows the probability that the group, exhibiting a character which comes out in the Megathymide, and which is dormant in the Skippers proper, | has left the main stem when this feature was in abeyance. That the forms are now relatively specialized is shown by the faded- out fork to the second anal of primaries. In specialization the branchlets of the radius have moved outwardly, keeping up their equidistant character. This is another link showing that the but- terfly wing had originally separate veins and a five-branched radius. Relics of this type are afforded by the Charaxinz in the Nympha- lids. All these butterfly families converge, and the discovery of a comprehensive type would make easy what now must be pieced out with much care and thought. At present the butterflies are all greatly specialized, but the generalized characters are still there in fragments and can be put together. ‘To these all the families contribute their quota. The Papilionides remain excluded, from their peculiarly placed third anal vein of the primary wing. As Ihave repeatedly said, grant- ing the homology of the fork attached to second anal vein of the Hes- periades, with the short, downwardly curved third anal vein of the Papilionides, the diphyletism would be shown in the fact that the third anal has joined the second anal in the one case and remained free in the other. How important this is may be inferred from the fact that the first position is held unchanged to the Tineides through immense series of forms in all conceivable conditions of specializa- tion, while that of the Papilionides is equally tenacious through a small group of genera of very diverse shape and appearance.

Family PIERIDID.

Wings entire, primaries broad trigonate, secondaries full, rounded, rarely with projections. Radius three to five-branched, cells closed, sometimes partially degenerate, again with slight backward projec-

1900. } GROTE—THE DESCENT OCF THE PIERIDS. 19

tions, indicating the vanished base to the median system. The two upper branches of the median system ascending the lower side of the radius in specialization. (The second branch follows suit in Phulia as well as in Pseudopontia.) M2 radial. The fork to the second anal vein of primaries is usually, perhaps always, present, rigid, very thin and transparent. Hind wings with the subcostal cell lost, the subcostal vein solid at base ; in some species, ¢. ¢., Hebomoia glaucippe, there is a narrow slit perceivable. I have therefore assumed that all traces of A1 have generally disappeared and lettered the Subcosta S. alone ; the fact that the radius is theo- retically furcate at base has no practical value here. The humeral spur (precostal spur) is usually present ; in some specialized forms, e. g., Nathalis, Colias, it has vanished ; in some subspecialized gen- era, Eurema, Zerene, Phedis, it has left a slight projection. Discal cell of secondaries closed. M2 radial. Two anal veins. As compared with the Lyczeni-Hesperiadz branch, the neuration shows more complex changes, the relations more netlike.

No characters of the wings, or, in fact, other, have been found by me by which a subdivision of the Pierididz could be effected. Where there are no gaps there are no categories. The generastand in various degrees of relationship toward each other; there is a tendency to form lines, culminating in a form with three radial branches, terminat- ing in one with the generalized number five. I have indicated these lines in a former communication to the American Philoso- phical Society. The phylogenetic lines come out more clearly in the holarctic fauna; as we approach the tropics the increased num- ber of forms render their separation quite difficult. Ihave no doubt I have often failed to give the true relationship; at least, I have tried to do this, to express by a lower position the fact that the form so placed displays here a neurational feature which a succeed- ing form has abandoned ; this is in principle what is intended by phylogeny.

A rough division into ‘‘ Whites’ and ‘*‘ Yellows ’’ from the pre- vailing colors of the wings has been attempted. Since these groups cannot be actually limited, they merit no Latin term. As often as I have tried to express their characteristics I have desisted, fearing to dwell upon exclusive marks. In Prerzs and Colas, or Hurymus, we have a typical expression of the two associations, which are pos- sibly really natural groups kept apart by general color.

Since the most specialized form I have found belongs to the im-

20 ’ GROTE—THE DESCENT OF THE PIERIDS. [Jan. 5,

mediate line of the typical genus Pierzs, | have commenced with the ‘‘ Whites ;’’ but since the lines probably really converge, it is not more correct than to begin with the ‘‘ Yellows.”

Typical ‘‘Whites.”’ (Pieridini. ) Gen. Phulia H.-S., 1867. Type: P. nymphula.

Phulia nymphula. Staudinger, /rzs, vil, Taf. i, Fig. 5 (neuration).

Radius three-branched ; Ar and #2 before the cross-vein, R3—5 to costa just before apex. Only the lower median branchlet from cross-vein ; d4/2 from lower side of radius not far from cross- vein; 73 a short furcation not far from apex. I cannot make out any trace of the fork to second anal; since this form is extraordinarily specialized by the ascension of the radius by dZ2 (paralleled in Pseudopontia), its absence might be expected, yet I have so often found traces of it in the Pierididz where I thought it absent, that it may be finally detected here. Secondaries oblong ovate, humeral spur pronounced, turning to base of wing ; first and second median together from upper angle of discal cell. No traces of subcostal cell.

From excellent figures of the neuration of species of Phudia given by Dr. Staudinger, /7zs, vii, Taf. i, Phulia nympha Staud. would be slightly more generalized by the propinquity of d/2 to the cross- vein (compare with Fig. 5 of nymphula). The species of Phulia, as well as the following Z7zfurcula, are Andean, flying at a height of 3000 to 5000 metres, department of La Paz, Bolivia.

Gen. Zrifurcula Staud. Type: T. huanaco.

Trifurcula huanaco. Staudinger, /rzs, vii, Taf. i, Fig. 7 (meuration).

Radius four-branched ; #2 advanced to opposite discal cross- vein; #3 a very short furcation. #1, ascending radius; JdZ/2 from cross-vein near to radius, so that the three veins are brought here close together, a feature which Staudinger considers charac- teristic. A very faint trace of fork to second anal. Hind wings shaped like Phufa; humeral spur straight, abbreviate; no trace of subcostal cell; JZ2 radial. _ More specialized than Pers by the advance of R2 to a point

1900. | GROTE—THE DESCENT OF THE PIERIDS. A

opposite cross-vein ; more generalized than Phulia by the four- branched radius and J7/2 still from cross-veins.

Gen. Mancipium Hiibn., 1806.

Type: M. brassicze.

Mancipium brassice. Grote, Mitt. a. d. Roem. Mus., 8, Vaf. i, Fig. 4 (neuration).

The validity of this genus seems to depend upon the fact, that in ’ my preparations the short fork, A3 of Pzervzs, has disappeared. Dr. Chapman informs me that in material examined by him the vein- let still shows itself. Zancipium is then a Preris with sometimes three-branched radius. It shows that there exists a: tendency in Pieris to lose R3; the individuals not reproducing the vein are specialized, those repeating it relatively generalized. The genus stands on a very uncertain footing and should probably be dropped ; it can only be used with any separate meaning for a part of the ma- terial of the typical species. Pontia daphdice, usually mixed with Pieris, may owe its similarity, I have ventured to urge, to converg- ence and belong really to the line Anthocharini. The generic title Mancipium covers the Preris of the future, not of the present day, since any /Pverzs discarding the short veinlet #3 would be- come thereby equally entitled to be referred to it. From the neura- tion, I cannot distinguish Ponta from Mancipium.

Gen. Preris Schrank, 18o1. cby pera r. rape.

Pieris rape. Grote, Mitt. a. d. Roem. Mus., 8, Taf. i, Fig. 5 (neuration).

Radius four-branched, 2 from above cell before cross-vein, #3 a very short fork before apices. Discal cross-vein shrunken ; Mx from radius, (4/2 radial, quite close to upper angle of the cell. Fork to second anal at base distinct. Hind wings with humeral spur turned outwardly; no subcostal cell, but the vein seems to show a cleavage, its remnant. Discal cross-vein also medially deteriorate ; 7/2 radial.

Pieris napi. Agrees closely with the preceding, but the second radial is ad-

vanced nearer the cross-vein (emulating Z7ifurcula). The fork #3 is still shorter, quite minute and I should expect it to be occa-

22 GROTE—THE DESCENT OF THE PIERIDS. [Jan. 5,

sionally absent, when the specimen would be referable to A/ancz- plum. PP. napi is thus more specialized than P. rape.

We have thus come somewhat rapidly from local butterflies living on the heights of the Andes to the typical genus of the holarctic Pierididz. It is clear from the retention of 72 above the discal cell, that the ancestors of Phudia have passed through the type of Pierts. Trifurcula seemingly represents an intermediate but lateral stage, in which 2 is advanced to opposite cross-vein, while JZ2 has not followed Aft and left the cell. The specializations are therefore probably independent and Pua has not passed through a Trifurcula stage.

The genus Pzerzs represents an advanced type of the four-branched radius, on the verge of becoming three-branched, but one in which the second radial has retained its original position above the cell. This type may be used for comparison with the other genera, some of which have evidently attained, through convergence, the same condition ; while on the passage of the five-branched to the three- branched form the stage of Pzerzs is more or less nearly repeated in different phyletic lines. For instance, in the ‘‘